The platform economy is rapidly colliding with AI, and while traditional prediction markets have exploded into the mainstream, a brand new asset class is emerging called information markets.
Now instead of just placing gambling bets, communities can trade highly specialized knowledge, and these networks are throwing out centralized gatekeepers entirely, allowing creators to monetize their niche audiences while using AI algorithms as the.
Neutral judge to settle the trades.
Well joining us to break down how AI is reshaping the creator economy is Ben Fielding, co-founder & CEO of Gensyn.
Thank you so much for joining us when it comes to prediction markets that we know some of the names out there as well as a poly market as well as how they have become household names.
But when it comes to your space and what you're doing, tell us about.
Information markets and what does that mean?
Sure, information markets are different from prediction markets in the fact that they allow anyone to create a market.
So in traditional prediction markets on the existing platforms, the platforms themselves create the markets and people can just trade the information that they have in those markets, but that's only one directional.
If people have information, they can monetize it through trades, but they can't ask a question of the world, whereas in information markets anyone can. a market exist and ask that question to the world and solicit information back and so it creates this bidirectional trade of information that we believe is essential for the future of AI.
That information needs to be in the public domain.
If you look at the existing large hyperscale AI companies, they're collecting information themselves and using it to monetize through their models, but there needs to be an open flow of information that can exist for open models, and that's where information markets come in.
Yeah, and before we move away from this, I do want to ask you the difference between centralized versus decentralized when it comes to information.
Sure, so in the centralized world, somebody is in control of that information, wherever it goes, that one entity who creates the market or creates the platform controls what happens to the information.
They could monetize that information in various ways, etc.
They might be altruistic in the first instance, but you never know what could happen in the future.
In a centralized world, that person is always in control.
In a decentralized world, you don't have that control.
That information is in the public and it's controlled by the users of the platform itself and so it maintains in the public forever.
One of the kind of core principles of blockchain and decentralized tech in general is that once it's created, it can't be made centralized again.
It's forever in the kind of decentralized world and owned by everybody who participates.
And because we're also talking about AI here on Wall Street, we pay attention to the Technology, its effect on the labor market and also in terms of an investment theme, how much hyper scalers are spending on caps and the return on investment.
But of course when it comes to this technology, some may say doesn't AI hallucinate.
What would you say to them?
Sure, I would say AI does hallucinate.
I think within prediction markets and information markets we have an Oracle problem essentially we need to know what actually happened with an event when a trade has been put on that.
Don't have a good way of solving that right now you have different ways the prediction markets will either use a centralized company, they just decide, or they'll use some kind of voting system over lots of people, they use human consensus, things like that.
Ultimately all of these systems are gameable.
You can sway public opinion in various ways.
If you weight that with financial stake, then you can sway that with enormous amounts of money.
What we believe is committing to an intelligence which is fixed before the event. is the cleanest way that you can possibly solve that Oracle problem because that intelligence cannot be swayed between the time the market is created and the event.
It can't be changed.
It is a singular commitment to a specific state of that artificial intelligence that will not change in any way, and anybody can rerun that intelligence and see that it did actually make that decision.
So it's an infallible judge as long as the people who trade in the market accept that it is fixed from the point the market is created. and many of us know about the creator economy, but I want to get your take on what you call the next creator economy.
So incentivizing communities to trade information actually work you break this down for us so if you think about somebody who has a large following, somebody who has a niche community.
Maybe they've got 100,000 followers in their community, they push out sub-stack articles, they talk to their community over X, etc.
That person's curated that community around a specific interest.
That specific interest is their niche to the community and that community has information about that interest because they're interested in it.
In the past, the way that those creators would monetize their community would be advertising to them.
They would use that niche set of information as a way to target ads to the community, but that's quite extractive.
It means that community is being bombarded by ads maybe they don't want to see.
In the information market future, they can monetize that community by posing questions to the community and saying you are knowledgeable about this, you're interested in this topic.
Now trade your knowledge and you should be more knowledgeable than everybody else because it's your niche interest.
What that does is it allows creators to bring their community into the discussion instead of them just making predictions on their live streams and things and their community listening.
They're actually able to pose a question and the community is part of the answer.
We know from prediction market research and from financial market research in general that wisdom of the crowd is true.
Soon as you get people putting stake behind what they believe, you get more truth than you would get from any single person.
And so this allows that creator to solicit truth from all of their followers.
Their followers become part of the conversation, and in doing that they monetize the actual trades that happen within the system, so they're able to make revenue from their users without extracting from them by blasting ads at them or something.
And then, while I have you here, I know you've been in the ML space for over a decade now, so I do want to get your take on where you think regulation should fall when it comes to both prediction as well as information markets.
Sure, so ultimately I believe the technology underneath information markets, so particularly decentralized information markets where anyone can create a market, anyone can trade in a market, and it can be settled by an independent AI model that is foundational technology.
It isn't controlled by any one person once it's created.
And so I view that in the same way I view something like encryption or if you go into the crypto space something like Uniswap that allows anybody to create pools over any asset, foundational technology that can be applied in applications for information markets so you can have any number of websites which allow you to trade in niche sets of information markets, broad sets of information markets.
Once the application is created, then there needs to be a regulatory discussion so somebody can say, OK, this application is created over markets that we don't like.
If we don't like that from a social governance perspective.
We can regulate that application, but what we shouldn't do very clearly is regulate the technology underneath it the same way we wouldn't regulate encryption in order to solve various problems in the world, because it has great benefit to things like banking.
Like the whole banking system would collapse if we regulated encryption, but as a foundational technology we need it to exist for huge benefits.
And then finally, before I let you go, what is the long-term vision for Jensen?
Sure, so the long term vision of Jensen is to essentially just open AI to everybody.
So we believe in scaling AI and machine learning as far as it can possibly go.
The way that it's been scaled so far by the hyperscale is very centralized.
The value of the information that is created by the users is captured by that company, and we think for a technology this important, that isn't the correct way to go.
We think if you take the model of the internet, for example, the way the internet grew as a technology, it allowed much more to happen because anybody was able to access it, anybody was able to use it, and it didn't have these gates over the top of it.
AI is progressing right now.
It does have gates, and we don't think that's correct.
There is a way to build it similarly to the internet in the open, owned by the people who contribute to it, and that's what Jensen builds fundamentally.
We build the infrastructure to allow that to happen.
Delphi, our information market platform, is an example of what you can do once you have that infrastructure, but there are many more things that can be built on top of it, crucially owned by the users, not owned by singular companies.
Well Ben, we will have to leave it there for today, but thank you so much for joining us live here at the New York Stock Exchange.
Thank you for having me.
Thank you.
